From Molecule to Metaphor : A Neural Theory of Language

Teaching With the Book

The book is obviously not a text, but it is already being used successfully in a variety of undergraduate and graduate courses. In fact, the current version of the book is the result of several years of my teaching (often with George Lakoff or Srini Narayanan) a UC Berkeley class called "The Neural Basis of Thought and Language". This is an upper division UG class, cross listed in Cognitive Science, Computer Science, and Linguistics. Our course involves the use of several complex systems and is not portable in its current form. One surprising outcome of the most recent offering was a complete graphical summary of each chapter, done by John Torous in studying for the final.

This is now part of this site as Reader's Roadmap.

The paperback version of the book is scheduled for January 2008 and should be a bit cheaper.

The MIT Press in-house Textbook Rep is Michelle Pullano (mpullano@mit.edu).

I would love to hear about other (good or otherwise) teaching experiences.

UC Berkeley Cog110/CS182/Ling109

Jerome Feldman

This is a one semester upper division UG class cross listed in Cognitive Science, Computer Science, and Linguistics. The material in the book is supplemented by a number of online readings.

The syllabus provides an overview of the course. The lecture schedule follows the book fairly closely and is available on request.

There are weekly assignments, many of which involve using NTL systems and also some coding exercises for the Computer Science students.

UC Berkeley Cog101/Ling105

George Lakoff

This is a large introductory class on Cognitive Science and Linguistics. George did not assign the M2M book in advance, but did later suggest that the students all buy it. One of the course assistants, Zhenya Antic, reports:

The chapters assigned were 1-5, 7, 9, 11, 13, 17, 18, 20. We went over the earlier chapters in section, but I don't know how much of that students actually "digested", since there was no homework on that.

~~~

I gave two guest lectures, covering roughly the material in Chapters 21 and 23-25. Much of the material in the rest of the book is already part of this class. The following question was on the (take home) final exam:

Consider the sentence:

"Karen was shoving the red cloth into the jar."

Suppose we would like to have the computer understand this sentence (supposing that this understanding will be somewhat like what our understanding of it is). Using chapter 11, 13 and 18 of From Molecule to Metaphor and your notes from Professor Feldman's lectures, describe what would be required for such understanding.

Questions you might want to consider:

What is understanding according to NTL?
What kinds of structures are required in order to understand the sentence above (for example, which image schemas)?
What are the bindings between different image schemas (which elements are mapped onto each other)?
What is specified in the sentence?
Which structures are evoked (i.e. are implicit)?
Note the form of the verb; what kind of structure and in which configuration would be required for it?

Portland State Sp436/536

David Ritchie says:

The seminar was quite successful considering that it was fully on-line. I only assigned the first section of your book - on the biology of cognition. With only a couple of exceptions all of the students, undergraduate as well as graduate found the explanations clear and easy to follow, and very useful for understanding the concept of embodiment (which was new to all but a handful). With the exception of the two or three whose science backgrounds were not up to the biology, all the other students felt that the first part of your book should be assigned in future versions of the seminar. Three or four of the students expressed the intention to continue on and read the portions on modeling, either over spring break or later. A couple of students wanted even more detail on the biological material. I accompanied the book with some selections from Andy Clark (Being There) and closed the quarter out with Ray Gibbs' excellent 2006 survey of the experimental literature, and the response to this sequencing was very positive.

I may or may not assign the chapters on modeling in future courses. I must confess I'm not at all convinced by the computer modeling approach, but I do think you explain it about as clearly as anything I've read.

After re-reading your book more closely, I regretted not having assigned more of the book in my seminar. Next time I teach the seminar, I probably will. There is a lot of great stuff in the chapters I did not assign.

UC San Diego Cogsci 253 Semantics and Cognition

Gilles Fauconnier

Subject: Emergence and creativity in meaning construction

This class will be a seminar, with discussion of major recent work, guest speakers, and presentations by participants.

Its general theme is the radical contribution of cognitive science > to the empirical and theoretical study of meaning, starting with analogy, metaphor, and mental spaces in the late 1970's, and expanding during the last ten years into a broader framework of conceptual mappings, integration, and compression, with wide applicability to many products of the human mind.

We will look at four major books written in the last 2 years that deal with these issues from different perspectives (neuro,
linguistic, computational, discourse, art)

Gilles Fauconnier says:

The course you noticed (Cogsci 253) was a seminar, and your book was one of several recommended readings, but we didn't use it directly as a textbook.

I believe however that it would be an excellent textbook for our undergrad core class on language and meaning (Cogsci 101C) and also for electives within cognitive science that bear on semantics or analogy. I will try that next year and let you know how it goes.

Cognitive Semantics

Wilhelm.Oppenrieder@germanistik.uni-muenchen.de

Die kognitive Semantik befasst sich mit Bedeutungsphaenomenen unter dem Gesichtspunkt ihrer Verankerung in den geistigen Faehigkeiten des Menschen (im Unterschied zur systemorientierten strukturellen Semantik oder der mengentheoretisch fundierten formalen Semantik). Grundlage sind Ueberlegungen, die diese geistigen Faehigkeiten an das koerperliche Substrat, die spezifischen biologischen Eigenschaften des Menschen binden. Insbesondere Vorstellungen von prototypisch strukturierten und daher "unexakten" Bedeutungen aus diesem Umkreis sind in den vergangenen zwei Jahrzehnten auch allgemeiner populaer geworden. Die verschiedenen Ansaetze gehen jedoch weit ueber dieses bekannte "prototypische" Element einer kognitiven Semantik hinaus und erfassen neben der Wortsemantik auch die kompositionellen Aspekte der Bedeutung. Zudem wird durch die so genannten konnektionistischen Ansaetze eine exaktere Erfassung dieser Grundueberlegungen und ihre Modellierung durch Computersimulationen versucht. Im Seminar wird es darum gehen, die angedeuteten Grundueberlegungen und ihre Spezifizierung in verschiedenen Ansaetzen genauer kennenzulernen.

Literatur: Geeraerts, Dirk (Hg.) 2006. Cognitive Linguistics: Basic Readings. Berlin, New York. Feldman, Jerome A. 2006. From Molecule to Metaphor. A Neural Theory of Language. Cambridge/Mass., London.

< I will also ask him for feedback >